which consisted of buyers’ surplus V - P and sellers’ surplus P - cost, could be
maximized. If a buyer decided to resell the permit, the price had to cover both the cost of
not using the road (i.e. the value) and the cost of purchase of the permit. Any surplus or
deficit was added to the subject’s income and paid at the end of the experiment.
Trading occurred in the auction market when buyers (sellers) could submit bids (ask).
The system only allowed quotes which improved the current proposed trade, i.e. higher
bids or lower asks. Every trade was based on the current bid and ask price, and every
buyer (seller) could accept an ask (bid) and the transaction was executed immediately. A
transaction immediately removed both the standing market bid and the ask. Subjects had
full access to the traded prices for that period but not to the previous periods. Any surplus
or deficit incurred was announced at the end of each period.
In the final stage, the drivers had to decide whether to drive or not to drive. If the
decision was not to drive, the driver earned zero profit at this stage. If s/he decided to
drive, s/he was paid according to his/her valuation of using the road.
The experiment was conducted in the experimental lab in the School of Social Science,
Universiti Sains Malaysia. Upon entering the lab, the subjects were provided with the
instructions and rules of the experiment. The experimenter then explained and gave
exercises to the subjects. The experiment started after all the exercises were answered
correctly by the subjects. The experiment was programmed and conducted with z-Tree
experimental software (Fischbacher (2007)).
3. Theoretical predictions
The theoretical model used to predict the results of the experiment was Bayesian Game
Against Nature (BGAN), which derives from the Bertrand Perspective from Friedman
(1984). The experimental predictions are based on three observables as follows:
Allocative Efficiency